Hitomi Constraints on the 3.5 keV Line in the Perseus Galaxy Cluster The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation Aharonian, F. A. et al. “Hitomi Constraints on the 3.5 keV Line in the Perseus Galaxy Cluster.” The Astrophysical Journal 837.1 (2017): L15. © 2017 The American Astronomical Society As Published http://dx.doi.org/10.3847/2041-8213/aa61fa Publisher IOP Publishing Version Final published version Citable link http://hdl.handle.net/1721.1/110044 Terms of Use Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. The Astrophysical Journal Letters, 837:L15 (9pp), 2017 March 1 https://doi.org/10.3847/2041-8213/aa61fa © 2017. The American Astronomical Society. All rights reserved. Hitomi Constraints on the 3.5 keV Line in the Perseus Galaxy Cluster Hitomi Collaboration* (See the end matter for the full author and affiliation lists.) Received 2016 July 25; revised 2017 February 6; accepted 2017 February 19; published 2017 March 3 Abstract High-resolution X-ray spectroscopy with Hitomi was expected to resolve the origin of the faint unidentified E » 3.5 keV emission line reported in several low-resolution studies of various massive systems, such as galaxies and clusters, including the Perseus cluster. We have analyzed the Hitomi first-light observation of the Perseus cluster. The emission line expected for Perseus based on the XMM-Newton signal from the large cluster sample under the dark matter decay scenario is too faint to be detectable in the Hitomi data. However, the previously reported 3.5 keV flux from Perseus was anomalously high compared to the sample-based prediction. We find no unidentified line at the reported high flux level. Taking into account the XMM measurement uncertainties for this region, the inconsistency with Hitomi is at a 99% significance for a broad dark matter line and at 99.7% for a narrow line from the gas. We do not find anomalously high fluxes of the nearby faint K line or the Ar satellite line that were proposed as explanations for the earlier 3.5 keV detections. We do find a hint of a broad excess near the energies of high-n transitions of S XVI (E 3.44 keV rest-frame)—a possible signature of charge exchange in the molecular nebula and another proposed explanation for the unidentified line. While its energy is consistent with XMM pn detections, it is unlikely to explain the MOS signal. A confirmation of this interesting feature has to wait for a more sensitive observation with a future calorimeter experiment. Key words: dark matter – galaxies: clusters: individual (A426) – galaxies: clusters: intracluster medium – X-rays: galaxies: clusters 1. Introduction et al. 2016; stacked Suzaku clusters, Bulbul et al. 2016). Some of the nondetections were inconsistent with other detections The nature of dark matter (DM) is one of the fundamental under the decaying DM hypothesis (in which the line flux must unsolved problems in physics and astronomy. Direct particle be proportional to the projected DM mass), most significantly, searches in laboratories as well as searches for electromagnetic in a sample of galaxies (Anderson et al. 2015). We also note signal from celestial objects have been conducted with no here that the signal from Perseus reported by XMM, Chandra, unambiguous detection so far. X-ray observations of DM and Suzaku was higher than expected given the signal from the concentrations, such as galaxies and clusters, provide a probe rest of the cluster sample (B14). Astrophysical explanations of for a particular DM candidate, a sterile neutrino, which is the reported line, in addition to those considered by B14, have predicted to decay and emit an X-ray line (Dodelson & also been proposed; a critical review can be found in F16.An Widrow 1994; Abazajian et al. 2001). Early searches that fl ( extensive review of the recent observations is given, e.g., by provided upper limits on line ux and thus the particle decay ( ) ) ( Iakubovskyi 2015 . rate as a function of line energy which gives the particle As recognized in all previous studies, the above line detections mass) are reviewed, e.g., in Abazajian et al. (2012) and — ∼ ( ) were near the capability for CCD detectors for a 100 eV Boyarsky et al. 2012 . resolution, the line reported from clusters with a ∼1eVequivalent A possible detection was reported by Bulbul et al. (2014, ( ) ) fi width EW is a 1% bump above the continuum, easily affected hereafter B14 , who found an unidenti ed line at by errors in modeling the nearby atomic lines and in instrument E » 3.55 keV in the stacked spectrum of a large sample of calibration. A confirmation with a much better spectral resolution galaxy clusters using XMM-Newton EPIC MOS and pn. Within Hitomi ( ) was considered essential. , launched in 2016 February and their sample was the Perseus cluster its central region , whose lost in March (Takahashi et al. 2014, 2016) after having returned a signal was particularly strong. B14 also reported a detection groundbreaking spectrum of the Perseus cluster (Hitomi Colla- from Perseus with Chandra at the same energy. Boyarsky et al. ) ( ) XMM boration 2016, hereafter H16 , offered us such a possibility. We 2014 reported an detection in the outer region of present results from this data set below. Perseus. Urban et al. (2015) and Franse et al. (2016, We use h = 0.7, W=m 0.3, and W=L 0.7 cosmology. The hereafter F16) detected the line in several regions of Perseus ( ) Suzaku ( ) cluster heliocentric redshift average for member galaxies is with ; however, Tamura et al. 2015 did not detect it in 0.0179 (Strubble & Rood 1999), and the redshift in the CMB the same Suzaku data. The 3.5 keV line was also reported from frame is 0.01737, which gives dL = 75.4 Mpc and a scale of other objects, such as the Galactic Center (Boyarsky et al. ′ ( ) fi ) ( ) 21.2 kpc per 1 . We use the 68% 1s con dence level for 2015 and M31 Boyarsky et al. 2014 . Other sensitive searches errors unless stated otherwise. did not detect a significant line signal (e.g., from the Milky Way halo, Sekiya et al. 2016; Draco dwarf, Ruchayskiy 2. Data * Corresponding authors: M. Markevitch, C. Kilbourne, and T. Tamura fi Hitomi ([email protected], [email protected], tamura. The Perseus cluster was the rst-light target for , [email protected]) observed early in the instrument activation phase with the Soft 1 The Astrophysical Journal Letters, 837:L15 (9pp), 2017 March 1 Aharonian et al. X-Ray Spectrometer (SXS; Kelley et al. 2016). SXS is an array we do not attempt to account for PSF scattering in and out of of 35 calorimeter pixels with a 4.9 eV FWHM energy the FOV, and use the instrument response for an on-axis point resolution (H16), covering a 3¢´3¢ field of view (FOV) at source. We estimate the effect of this simplification on the the focus of a Soft X-Ray Telescope (SXT; Soong et al. 2014; model normalization to be ∼10%. Okajima et al. 2016). To maximize statistics, here we coadd the The uncertainty of the Be layer thickness in the GV window, 230 ks Perseus data set used in H16 and a later 45 ks pointing 10m m, corresponds to a ±2.5% uncertainty for the flux for a total exposure of 275 ks. The former data set is a at E = 3.5 keV. combination of observations 2 on 2016 February 24–25 and 3 A more insidious effect may be caused by uncertainty in on March 3–5, both pointed ∼1′ away from the cluster center, modeling the SXT effective area (Kurashima et al. 2016). The while the latter (observation 4 on March 6–7) is on-center. The SXT reflectivity around the Au M edges was measured on the earliest observation, observation 1, was pointed away from the ground and combined with values from Henke et al. (1993) for core and is not included. other energies. The ground measurements show ∼1% systema- For these observations, SXS was still protected from tic deviations from Henke, one of which is in the possible contaminants by the closed gate valve (GV) window. 3.43–3.68 keV interval above the Au M1 edge—at our energies It includes a Be filter that absorbs soft X-ray photons. At of interest. Given the finite accuracy of the ground measure- E = 3.5 keV, the GV window transmission is 1/4 of that in the ments, we consider the possibility that the Henke values are normal operation mode, yielding the number of photons more accurate. To quantify the effect of this uncertainty, below equivalent to about 70 ks of normal observations. we will derive some of the results using both the default area curve (which uses the Hitomi mirror measured reflectivities) 3. Analysis and one in which the Henke values were used above the Au M1 edge. Similar deviations may be seen at other Au M edges, but To fully utilize the SXS high energy resolution, accurate ( ) ( the next one M2 at 3.15 keV is well outside our interval of calibration of gain the conversion from the amplitude of the interest and we will not consider it. detected signal to photon energy) for each of its 35 pixels is essential. Unfortunately, the individual pixel gains were changing during the early part of the mission, and a 4. Results contemporaneous gain calibration for the SXS array as planned for later operations was not available.